Genetic linkage studies implicate a gene or genes at Xq27 in hereditary prostate cancer susceptibility (HPCX). The corresponding region spans 750 kb and includes five SPANX genes (SPANX-A1, -A2, -B, -C, and D), which encode proteins that are expressed in sperm nuclei and a variety of cancer cells. Each SPANX gene is embedded in a recently-formed segmental duplication (SD) up to 100 kb in size, resulting in extensive enrichment in long stretches of repeated DNA in this gene region. Due to their recent amplification, both SPANX coding and flanking sequences in the SDs are nearly identical throughout the SPANX-A/D cluster, which complicates sequence analysis of these genes by PCR-based methods in the search for mutations. However, we recently succeeded in performing such an analysis of the Xq27-linked SPANX genes from prostate cancer patients, using the transformation-associated recombination (TAR) technique, which makes it possible to directly isolate large genomic segments from complex genomes. This analysis revealed frequent gene deletion/duplication and homology-based sequence transfers involving SPANX genes at Xq27, suggesting that SD-mediated homologous recombination involving the SPANX genes might lead to increased genetic instability and possibly to a higher level of genetic diversity in SPANX genes in germ lines. The results of the analysis showed that no DNA sequence variation or genetic haplotype in the SPANX gene cluster was associated with susceptibility to prostate cancer. However it remains possible that Xq27-linked prostate cancer susceptibility is related to variation in the architecture of the SPANX-A/D gene cluster. We hypothesize that X-linked predisposition to prostate cancer is caused by SD-mediated genomic rearrangements at Xq27. It is well known that SDs mediate ectopic interactions between distal chromosomal sites leading to chromosomal rearrangements such as duplications, deletions, and inversions. Such SD-mediated rearrangements can alter expression of genes in the vicinity of the SD, resulting in cellular and/or phenotypic changes and pathological disease. The density of SDs in the SPANX gene cluster is unusually high, representing more than one third of the 750 kb genomic region, suggesting a likely hot spot for genomic rearrangements. Based on the structure and organization of duplicated segments at Xq27, a high frequency of deletions and duplications of SPANX genes and flanking DNA as well as inversions of SPANX-containing regions is predicted. Some of them may result in altered expression of one or more genes within or near the SPANX gene cluster. Therefore, future work will focus on identifying inversion(s) within the SPANX-A/D gene cluster in X-linked prostate cancer families, which could lead to malignancy. After more than two decades of investigation, human centromeres remain enigmatic and poorly understood. Some progress in this field was outlined after demonstration by Hunt Willards group that alpha satellite DNA (alphoid DNA), the primary DNA found in human centromeres, can induce the seeding of a kinetochore complex in human HT1080 cells. Several groups have confirmed this observation and reported the formation of Human Artificial Chromosomes (HACs) in human cells, using a transfection strategy that involved alpha satellite DNA. These HACs are maintained as single copy episomes (i.e., copy number of 1) in the nucleus and have a fully functional kinetochore. The development and detailed studies of HACs offer new approaches for: 1) elucidating the mechanisms for de novo centromere/kinetochore formation and its structural/functional organization, and 2) creating gene delivery vectors with potential therapeutic applications. Until recently, HACs were constructed from 50-100 kb alphoid DNA fragments identified in existing YAC or BAC libraries. As a rule, the complete DNA sequence of these fragments was unknown, which did not allow making a final conclusion regarding the structural requirements for de novo kinetochore formation. Following the establishment of our unit in NCI, we have focused part of our research on constructing synthetic alphoid DNA arrays with precisely-defined DNA sequence variations. For this purpose, we developed a novel method for construction of alphoid DNA arrays, CADA, exploiting in vivo recombination in yeast. This method represents a tool for mutational analysis of alphoid DNA, as each of the constructed arrays can be evaluated for its ability to form a HAC and if necessary can be modified. Knowledge of combinatory of DNA sequences initiating HAC formation from alphoid DNA substrates may provide unique information on the functional organization of the mammalian centromere. It may also accelerate the construction of a more sophisticated HAC-based system for gene delivery and expression. Among the main unsolved problems of the HAC system are: i) a low efficiency of de novo HAC formation, and ii) multimerization of the input DNA concomitant with a HAC formation, making it difficult to control gene copy number and the location of genes in a HAC. Optimization of the centromeric component of HAC constructs is likely to be key to ensuring an efficient seeding of a functional kinetochore, and may also influence the extent of DNA multimerization during HAC formation. Using CADA method, we have recently determined the minimal size of alphoid DNA array capable of forming a HAC and constructed a novel HAC to manipulate the epigenetic state of chromatin within an active kinetochore. The HAC has a dimeric alpha-satellite repeat containing one natural monomer with a CENP-B binding site, and one completely artificial synthetic monomer with the CENP-B box replaced by a tetracycline operator (tetO). This HAC exhibits normal kinetochore protein composition and mitotic stability. Targeting of several tetracycline repressor (tetR) fusions into the centromere had no effect on kinetochore function. However, altering the chromatin state to a more open configuration with the tTA transcriptional activator or to a more closed state with the tTS transcription silencer caused mis-segregation and loss of the HAC. tTS binding caused the loss of CENP-A, CENP-B, CENP-C and H3K4me2 from the centromere, accompanied by an accumulation of histone H3K9me3. In addition to providing the first clear demonstration that heterochromatin within the centromere is incompatible with kinetochore activity, the conditional centromere of the HAC opens a new spectrum of opportunities for the systematic manipulation of the histone code within the kinetochore, and definition of the full epigenetic signature of centromeric chromatin. The new HAC with a conditional centromere also has potential as a system for gene delivery and regulated gene expression in mammalian cells. To make possible a regulated gene expression in HACs, the tet-O containing HAC was transferred from human host cells into chicken DT40 cells exhibiting a high level of homologous recombination. Our recent studies show that the HAC with a conditional centromere has high mitotic stability and is maintained in a single copy in DT40 cells. To generate a HAC-based system for expression of mammalian genes, a unique lox-P was introduced into the HAC with a conditional centromere using a recombinational machinery of DT [summary truncated at 7800 characters]